The influence of green roofs on the rainwater management system in an urban, tropical and undeveloped environment

The influence of green roofs on the rainwater management system in an urban, tropical and undeveloped environment

Research Area

San Pedro Sula, Honduras

Herman van der Bent Bachelor Assignment

10-09-2009

Bachelor Assignment 10-09-2009

H.S. van der Bent University of Twente

Civil Engineering & Management h.s.vanderbent@student.utwente.nl Supervision by:

Arcitect A. Stassano

Techos Verdes, Plaza Comercial Bioclimática San Pedro Sula, Honduras

Supervision University of Twente:

ir. J.E. Avendano Castillo

II Bachelor Assignment

Preface

This research has been done for my bachelor thesis of the study civil Engineering at the University of Twente, The Netherlands.

From April the 25

to July the 25

I lived in San Pedro Sula, Honduras, and did my research at the Plaza Comercial Bioclimática. The place commanded by the Architect A. Stassano is a good example of what can be done with green roofs. In her effort to enhance the further implementation of green roofs in the environment of San Pedro Sula, A. Stassano has built several green roofs and has carried out research regarding different aspects of green roofs. Last year a Dutch student Jitta Meijer did a research on the temperature effects of green roofs and as a follow up I was able to do a research on the subject of green roofs and rainwater management.

The research has not always been easy. A long time without precipitation and a political situation which forced me to leave the country for a week, made me creative and I approached the research question from a special point of view. After three months I found important insights regarding the water management capabilities of green roofs and I’m happy that I’m able to contribute to the efforts of A. Stassano in enhancing the implementation of green roofs into the environment of San Pedro Sula.

My special thanks go to A. Stassano for the opportunity which she gave me to do the research and for her help in finishing the investigation.

My thanks also go to ir. J.E. Avendano Castillo for being helpful in the preparation and correction of the report and his help in the arrangement of a good home address in San Pedro Sula.

At last I would like to thank Isaac Medina for his help in the preparation of the test setup and Jose Lopez, the local gardener, for helping me out in different situations and the effort he put in trying to improve my Spanish.

This research and stay in Honduras was an unique experience and I’m glad I had the opportunity to do it.

Herman van der Bent

August 2009

III Bachelor Assignment

Summary

The main research question was if green roofs are a good design solution to improve the water management system in an urban, tropical and undeveloped environment. To answer this question the water management capacities of an inexpensive extensive green roof design have been examined in the tropical city of San Pedro Sula. Due to the fact of a short research period the investigation has been approached from a water balance point of view.

The boundary conditions tell us the total amount of water which can be handled by a green roof. For example the water handling capacity is about 32 mm of water for an extensive green roof with a 7.6 cm thick soil layer and the local used soil mix. This 32 mm of water is divided into three areas: The bottom capacity, the retention capacity and the drainage capacity. Out of these boundary conditions two main issues have been extracted.

1. The evaporation rate per day. This is the amount of water which leaves a green roof and therefore equals the retention capability when a new rain event approaches. The retention capability is limited by the fact that an aggressive irrigation strategy is used which waters the green roofs after one day without precipitation. The maximum amount of water which can be retained is therefore maximized by the evaporation rate of one day. This is about 5 mm of water. If a better plant type can be found which survives with a less frequent irrigation strategy the retention capacity can be a bit higher.

2. The second issue is the ability of green roofs to delay the discharge of water. This gives the sewer system more time to handle large amounts of precipitation. The boundary conditions maximize this drain capacity at 3.5 to 8 mm, depending on the size and slope of the roof. Half of this amount is delayed more than between 2 and 6 minutes depending on the slope of the roof. Other not examined configurations of green roofs leave space for different times of delay.

These rainwater management capacities have been compared to the precipitation data of ten years in San Pedro Sula. These data were organized in small, middle and heavy rain categories.

Small 0-20 mm/day 99 days per year Middle 20-40 mm/day 12 days per year Heavy 40 -200 mm/day 6 days per year

The influence of green roofs with their maximum of 5 mm retention and the ability to slow

down 2 to 4 mm for 2 to 6 minutes is a positive number, but certainly on the heavy rain events

the influence is limited. It should not be considered that green roofs can totally solve problems

with the discharge of precipitation. A well maintained sewer system stays a governing design

solution.

IV Bachelor Assignment

Table of Contents

Preface ... II Summary ... III Table of Contents ... IV Table of Figures ... V

1. Problem definition ... 1

1.1 Research objective ... 2

1.2 Research model ... 3

2. Context of the green roof concept ... 4

2.1 Functions of green roofs ... 4

2.2 Extensive & Intensive Green roofs ... 5

2.3 Design parameters ... 6

3. Research: Water handling capacities of green roofs... 7

3.1 Research methodology ... 7

3.2 System Approach ... 10

3.3 Boundary conditions ... 12

3.4 Retention ... 18

3.5 Run off & drainage ... 29

3.6 Conclusion water handling capacities ... 36

4. Research: Implementation ... 37

4.1 Characteristics of San Pedro Sula ... 37

4.2 Precipitation analyses of San Pedro Sula ... 40

4.3 Theoretical influence of green roofs ... 45

4.4 Large scale green roof engineering ... 48

5. Conclusions ... 50

5.1 Recommendations... 51

Appendix A – Visualization build up test setup ... i

Appendix B – Runoff prediction technique: Curve Number Method ... ii

Appendix C – Bibliography ... vi

V Bachelor Assignment

Table of Figures

Figure 1 Research model ... 3

Figure 2 Example extensive green roof ... 5

Figure 3 Example intensive green roof ... 5

Figure 4 Research model: Section field study ... 7

Figure 5 Test setup first research part ... 9

Figure 6 Test setup second research part ... 9

Figure 7 Example water content level in time ... 11

Figure 8 Table typical water contents of soil ... 12

Figure 9 Water content Boundary conditions ... 13

Figure 10 Table: Test setup results boundary conditions ... 14

Figure 11 Graph: Test setup results boundary conditions ... 15

Figure 12 Graph: Test setup results boundary conditions ... 17

Figure 13 Graph: annual sun radiation SPS ... 20

Figure 14 Graph: Annual maximum Evapotranspiration SPS ... 20

Figure 15 Crop coefficients ... 21

Figure 16 Graph Evapotranspiration SPS with crop coefficient factors ... 21

Figure 17 Average retention test setup ... 22

Figure 18 Evaporation rater Theory of Hargreaves combined with test setup results ... 23

Figure 19 Short time evapotranspiration model ... 24

Figure 20 Evapotranspiration after x-days without precipitation ... 25

Figure 21 Evapotranspiration after x-days without precipitation with irrigation strategy ... 26

Figure 22 Drain capacity green roofs ... 29

Figure 23 Effective drain capacity green roofs ... 29

Figure 24 Approximation delay in time in drainage ... 31

Figure 25 Drain water in roofs after x-minutes 5.1 cm soil - 2 degree slope ... 32

Figure 26 Drain water in roofs after x-minutes 7.6 cm soil - 2 degree slope ... 32

Figure 27 Drain water in roofs after x-minutes 5.1 cm soil - 20 degree slope ... 33

Figure 28 Drain water in roofs after x-minutes 7.6 cm soil - 20 degree slope ... 33

Figure 29 Approximation in delay in time in drainage ... 35

Figure 30 Boundary conditions test setup results ... 36

Figure 31 City map: San Pedro Sula ... 37

Figure 32 Map San Pedro Sula with rivers ... 38

Figure 33 Pictures sewer system San Pedro Sula ... 39

Figure 34 Pictures Water on streets after rain event ... 39

Figure 36 Comparison precipitation El ocotillo - Villas Mackay ... 40

Figure 35 Map location precipitation measurements ... 40

Figure 37 Yearly precipitation San Pedro Sula ... 41

Figure 38 Monthly precipitation San Pedro Sula ... 41

Figure 39 % days with precipitation in different categories ... 42

VI Bachelor Assignment

Figure 40 % amount of precipitation in different categories ... 42

Figure 41 Amounts of days with precipitation per category ... 43

Figure 42 Amount of precipitation per category ... 43

Figure 43 Appearance heavy rain events ... 44

Figure 44 Influence green roofs on small rain events ... 45

Figure 45 Influence green roofs on middle rain events ... 46

Figure 46 Influence green roofs on heavy rain events ... 46

Figure 47 Test setup (van Woerd, 2005) ... 47

Figure 48 Large scale green roof ... 48

Figure 49 Green roof layer one: Metal sheet ... i

Figure 50 Green roof layer two: Plastic layer ... i

Figure 51 Green roof layer four: Soil 5.1 & 7.6 cm ... i

Figure 52 Green roof layer three: Geo-textile ... i

Figure 53 Green roof - rainwater measurement system... i

Figure 54 Green roof layer five: Addition of grass ... i

Figure 55 Example curve number method ... ii

Figure 56 Typical curve numbers... iii

Figure 58 Green roof curve number ... iv

Figure 57 Table AMC- CN... iv

Figure 59 Pictures permeable pavements ... v

1 Bachelor Assignment

1. Problem definition

Rain water management in an urban environment always is an issue which needs special attention. Rainwater is collected very quickly due to the fact that cities have lots of impermeable surfaces. With heavy precipitation this could lead to flooding of the streets with economical and social damage as a result.

This issue includes the place of research, San Pedro Sula, Honduras. Several news articles mentioned high amounts of precipitation in this city with a tropical climate. An example occurred on 24 October 2008. Large rainfalls caused the surroundings of San Pedro Sula to flood. 22.000 people had to be evacuated and 22 people lost their lives (Children International, 2008). To solve this water management problem several research directions are available. Some directions are to examine the flow of water in rivers, to do more research on the sewer systems or to explore possibilities to store rainwater in safe areas.

In Honduras an architect, A. Stassano, is convinced of the sustainable building theme. She is exploring the possibilities of sustainable building solutions to help the city forwards. Introducing green roofs as a profitable building concept is one of her issues.

Several studies mention the rainwater handling capabilities of green roofs. (Berghage R.D, 2009) (Mentens, et al., 2003) (Mentens, et al., 2005) Most of these studies did take place in developed countries with relative high financial means. This delivered data of the quality of green roofs and delivered some green roof concepts with high quality standards. The effect of green roofs on the rain water management system is found to be positive in these studies.

The two main interesting capabilities of green roofs to handle rainwater are the capability to

retain a part of the rainwater and the second, to delay an amount of rainwater. Green roof are

able to store rainwater in the soil layer and evaporate it back into the air. This retention factor

will decrease the discharge of rainwater to the sewer system and therefore improve its

performance. The capability of green roofs to delay an amount of water can help the sewer

system to handle heavy rain events. The sewer system has more time to handle the rainwater,

due to the slow release of rain water out of the green roofs. (Mentens, et al., 2003) (Van Woerd,

2005)

2 Bachelor Assignment 1.1 Research objective

This research has been done to find out if the water handling capacities of green roofs also contribute to a better rainwater management system in an urban, tropical and undeveloped environment.

Objective:

The main objective is formulated as follow:

To determine if green roofs are a good design solution to improve the rainwater management system in an urban, tropical and undeveloped environment.

Terms explained:

Green roofs:

A roof design with a soil layer containing plants placed on top of the roof. The concept is explained in detail in chapter 2.

Rainwater management system:

The system which takes care of the discharges of rainwater from a city level to a river based level.

Urban environment:

In cities with high amounts of paved surfaces the rainwater is drained rapidly during a rain event.

Tropical environment:

Tropical environments have a typical distribution of rain events. The rain events in a tropical climate normally appear with high amounts of water and therefore their impact on the city environment is large. Therefore specific research is done on the precipitation characteristics of San Pedro Sula.

Undeveloped environment:

A large part of the city of San Pedro Sula is not highly developed. Therefore an inexpensive and simple green roofs design should be used. The financial means are scarce and need to be used wisely. This is considered in the designed test setup. The design is simple and relatively inexpensive.

Research questions

To fulfill the research objective three main questions have been answered:

1. What is the theoretical context of the green roofs concept?

2. How do the water handling capacities of green roofs work in a tropical and undeveloped environment?

3. What is the influence of green roofs on the urban water management system?

(San Pedro Sula)

These questions are answered respectively in chapter 2, 3 and 4.

3 Bachelor Assignment 1.2 Research model

The research has the following setup to answer the main question if green roofs are a good design solution to improve the water management system in an urban, tropical and undeveloped environment. A test setup is used in the research phase on the water handling capacities. Because of the strong relationship with theory the test setup results are presented together with the describing theory.

Problem definition with main question

Theoretical context of the green roof concept

Research on the water handling

capacities of green roofs

Implementation Green roofs San Pedro Sula

Characteristics SPS

&

Precipitation research

Research results

Theoretical influence green

roofs on water management

system Chapter 1

Chapter 2

Chapter 3

Chapter 4

Conclusion Chapter 5

Run off & drainage System analyses

Retention Boundary

conditions

Theory

Test setup results

Theory

Test setup results

Theory

Test setup results

Figure 1 Research model

4 Bachelor Assignment

2. Context of the green roof concept

The green roof engineering concept uses the idea of using available space in the city as good as possible. The areas in the city we use unprofitable are the roofs on our buildings. The general idea is to place a substrate layer on top of roofs. It is possible for plants to grow in this substrate layer.

2.1 Functions of green roofs

A green roof is able to fulfill several functions at the same time.

The main functions of a green roof are:

1. To be the upper layer of a structural building and shield off an inside space.

2. To improve insulating capacities of a roof for better inside temperature conditions. It saves energy on cooling or heating and therefore creates a financial benefit for the owner. (Barrio, 1997) (Meijer, 2009)

3. To behave as a sun energy consumer. Plants use sun energy and reflect less heat into the air. This could lower outside temperatures. This effect is known as the urban-heat-island effect. (Meijer, 2009)

4. To act as a water retaining buffer. Decreasing the amount of impermeable surfaces in a city improves the water management performance. (Mentens, et al., 2005)

5. To improve the quality of esthetics in a city. A green environment will give people a good feeling. (Dunnet, et al., 2008)

All the different functions demand different designs of green roofs. From a water retaining and insulation point of view it is good for a green roof to have a thick substrate layer. On the other hand the weight should be as low as possible for just functioning as an upper layer of a construction, because in this way a smaller and more cost effective construction can be used.

Mostly the demands of the owner of the building will determine the dimensions of a green roof

and thereby how well it functions in each of the categories.

5 Bachelor Assignment

2.2 Extensive & Intensive Green roofs

To be able to classify different types of green roofs a distinction is made between extensive green roofs and intensive green roofs. Intensive green roofs have a thick substrate layer so bigger plants will be able to grow in it. The roof acts like a garden and needs maintenance. An extensive green roof needs far less maintenance and could be a sedum or a thin soil layer with grass or groundcovers. (Mentens, et al., 2003)

Intensive green roofs

An intensive green roof acts as a real garden on top of a roof and has therefore more opportunities for esthetic appearances. Also due to the thicker soil layer an intensive green roof has slightly better performances on acting as an insulation layer and as a buffer for precipitation. Nevertheless it has a large disadvantage on just functioning as the structural top layer of a building. A well designed structural plan is needed to support an intensive green roof.

This will increase the costs of a building and also can be inconvenient when placing a green roof on existing buildings. Also can be considered that maintaining an intensive green roof will take its necessary costs.

Extensive green roofs

In contrary with intensive green roofs, extensive green roofs do not acquire an extended structural plan. They can be placed on most roofs, because of a relatively low weight. The costs for an extensive green roof will be lower compared to intensive green roofs and also less maintenance is required. The esthetics of an extensive green roof have boundaries because of the thickness of the soil layer. It is not possible to plant trees, but different types of sedum, grass, ground covers and small plants survive on extensive green roofs.

Figure 3 Example intensive green roof Figure 2 Example extensive green roof

6 Bachelor Assignment 2.3 Design parameters

Green roofs can appear in a wide variety. It is up to the owner and the architect to choose between different options. Options to be chosen out are: the slope of the roof, the orientation of the roof, the buildup of the soil, the thickness of the soil layer, the presence of a drain layer, the used plant type and the installed irrigation system. All these parameters have an influence on the water handling capabilities of green roofs.

Besides the parameters which can be influenced by the design, also the period of the year with different weather conditions and the intensity of a rain event influence the capabilities of a green roof to handle precipitation.

The design parameters which influence the water handling capacities of green roofs are discussed in the chapters were the parameters have an influence. This can be in the area of the boundary conditions, the retention theory or the drainage theory.

Construction layers

A green roof consists of several construction layers. A typical build up of a green roof consists out of the following layers:

- Construction layer - Insulation layer - Waterproof layer - Root membrane - Drainage layer - Growing medium - Plants

- Drainage system

A full description of these construction layers can be found in the report: (Meijer, 2009)

The insulation layer, the drainage layer and a drainage system appear in sophisticated green

roof designs. The architect A. Stassano has been doing some tests with the layers on the Plaza

Comercial Bioclimática, but for cost effective reasons these were not added to the test setup

used for the research on the water handling capacities. (Chapter 3)

7 Bachelor Assignment

3. Research: Water handling capacities of green roofs

The water handling capacities of green roofs are analyzed to answer the main question if green roofs can improve the rainwater management system in an urban, tropical and undeveloped environment.

3.1 Research methodology

Research on the water handling

capacities of green roofs

Run off & drainage System analyses

Retention Boundary

conditions

Theory

Test setup results

Theory

Test setup results

Theory

Test setup results

Figure 4 Research model: Section field study

The analyses of the water handling capacities of green roofs start in the next paragraph with a system approach on the water content balance of a green roof. After this system approach the water handling capacities are further limited by the boundary conditions. The theory and the test setup results are presented in the same paragraph. This makes it easier to explain.

After the boundary conditions, the retention of rainwater is discussed. From a balance point of view the evapotranspiration in time is responsible for the capacity of green roofs to retain water. In this chapter a model for the estimation of the evapotranspiration is discussed and results of evaporation rates of the test setup are evaluated. The irrigation strategy appears to have a big influence on the retention capacity of green roofs. Also on this subject the theory and the test setup results are presented in the same paragraph.

The capacity of green roofs to delay a certain part of rainwater drained to the sewer system is

discussed. Boundary conditions give a first limit on how much water can be delayed and

drainage rates give an idea of time of delay. The slope has a big influence on the drainage

velocity and thereby on the effectiveness of green roofs to delay rainwater in time.

8 Bachelor Assignment

That makes the setup of the discussion of the water handling capacities as follow:

1. System approach: Gives the water balances 2. Boundary conditions: Gives boundaries of capacities

3. Retention: Given by the evaporation rate in time 4. Runoff and drainage: Gives time delays in run off

The different design parameters which have an influence on the three describing factors of green roof (the boundary conditions, the evaporation rate and the drainage) are discussed in the report if they have an influence on the specific water handling capacity. For the different aspect the design parameters are discussed as follow:

1. The boundary conditions

(Soil thickness, soil build up, plant type, slope) 2. The retention

(Plant type, slope, orientation, irrigation, year period) 3. The runoff and drainage

(Slope, soil build up, drainage layer, intensity rain event) Research constrains

A test setup was made to deliver good data of the water handling capacities of green roofs.

Some decisions had to be made in order to build a good setup which was able to deliver sufficient data, but for a reasonable price.

The first decision had to be made on the build up and proportions of the test green roofs.

Because the research question is focused on a research area with an undeveloped environment the choice for an inexpensive extensive green roof was made. The chosen build up for the test green roofs is visualized with pictures in Appendix A.

The second decision was made on the different design parameters which were closely examined. In consultation with the architect A. Stassano was decided to make a more detailed investigating of the influence of the soil thickness and the slope of the roof on the water handling capacities of green roofs. The influence of a drainage layer is not investigated because of a limited amount of available building material and due to the fact that a drainage layer will increase the costs of a green roof and therefore it is less easy to implement in an undeveloped environment. The used soil type is decided to be the soil mix the architect uses for green roofs.

This is a mixture of soil and gravel. This soil layer is planted with a grass layer so it will not take a

lot of time to fully cover the roof.

9 Bachelor Assignment Test setup

Out of the research constrains the decision was made to make four equally sized green roofs of one square meter. The decision was made in comprehension with the supervising architect.

Two of the green roofs had a soil thickness of 5.1 cm and two green roofs had a soil thickness of 7.6 cm. These sizes were chosen to cover a wide area of extensive green roofs.

The axe between the four roofs was placed in a east west configuration. This assured that the sun had an equal amount of radiation on every green roof.

During the first month the test roofs were placed under an angle of 2 degrees, so the first month there was an equal pare of roofs. Of as well the equal pare of 5.1 cm roofs and the equal pare of 7.6 cm roofs the data were compared, so an estimation could be made of the differences between two supposedly the same green roofs. The differences in research results between the two sets of roof were on average 2-3%, with a standard deviation of another 3-4%.

These numbers are reasonable. A full comparison is made in the data report.

After a month of research the roofs were placed under a slope to examine the influence of the slope on the water handling capacities of green roofs.

Tests

The data of the tests done on the green roofs are noted in the data report. De tests on the evapotranspiration part were done by adding water in a saturating level. The difference between input and output of water is the evapotranspiration (Eq 5). The tests done on the drainage research part were done by measuring the ouput of water after saturation of the green roof. After every certain time period (minutes) the output was measured. Then the drainage graphs were made and compared with an aproximation line. The graphs and tables can be found in the data report. Concluding data is presented in the upcoming paragraphs.

Figure 5 Test setup first research part Figure 6 Test setup second research part

10 Bachelor Assignment 3.2 System Approach

In the system approach the water balances for long time and short time periods are given.

These are the setup for the investigation and give a general understanding of the water handling capacities.

Water Balance - long time period

For longer time periods (months – years) the water content of green roofs is a balance between water going in and water going out. From time to time the water content can differ, but on a large time scale it has no net storage. The water balance for green roofs on a long time scale looks as follows:

in out

Water  Water (Eq. 1)

Water going in a green roof is always Precipitation or Irrigation. The water going out is always the sum of Runoff or Evapotranspiration. (Mentens, et al., 2003) There is no other way for water to enter or leave a green roof. Therefore the water balance can be written as follow:

I       t P t R t ET  t (Eq. 2) With:

I  t Irrigation during the time period P  t Precipitation during the time period R  t Run off during the time period

ET  t Evapotranspiration during the time period Water Balance - Short time period

For short time periods (hours – days) the capability of green roofs to store water cannot be neglected. Two types of occurrence can be examined. One during a rain event (or irrigation) and one in the absent of rain events. The balance for a rain event is as follow:

( )

P      t R t W

   t ET  t (Eq. 3) P  t Precipitation during time period

R  t Runoff during time period W

t

  Change water content level due to precipitation during time period (  ET  t ) Evapotranspiration during time period

(Considered to be neglectable during rain events) Therefore the change in water content level after a rain event is:

( )

W

t P R t

     (Eq. 4)

11 Bachelor Assignment

The other event for short time periods is in the absence of rain events. During these time periods only Evapotranspiration will take place. The water which leaves the roof comes directly out of the water storage. Therefore change in water content level is:

W

t ET t

    _{(Eq. 5)}

ET t

  Evapotranspiration during time period W

t

  Change water content level due to evapotranspiration during time period Water content level in time

The water content of a green roof is influenced by the two balances mentioned above: one during rain events (or irrigation) and one in the absence of rain events.

( )

W

t P R t

     &    W

t ET  t (Eq. 4 & Eq.5) The water content level in time can then be described by the equation:

( )

W t P R t ET t

       (Eq. 6)

To visualize what is happening with the water content in time an example of the progress in time is presented in the next graph. Notify certain limits of the possible water content of a green roof.

Figure 7 Example water content level in time

Explanation:

Day 1-5 no precipitation, evaporation and transpiration (no irrigation)

Day 6 heavy rainfall, water content high, after rain back to drainage maximum Day 7-9 no precipitation, evaporation and transpiration

Day 9 no precipitation, irrigation

Day 10-12 no precipitation, evaporation and transpiration

1 2 3 4 5 6 7 8 9 10 11 12

Water content

time (in days)

Example water content level in time

volumetric maximum drainage maximum plant minimum water content level

12 Bachelor Assignment 3.3 Boundary conditions

The water contents mentioned in the “example water content in time” always behaves between certain limits. These limits are different for different types of green roofs.

Volumetric maximum

The maximum water content level appears if a green roof is fully saturated. In nature this will never occur, because of the possibility to drain water at the bottom. The maximum water content of a green roof is determined by its soil buildup. The volumetric maximum per square meter is dependent from the thickness of the total soil layer and its porosity. The porosity differs with different types of soil. (Budhu, 2000)

max

*

W  H n

(Eq. 7)

With:

W

Maximum water content level in mm H Thickness of the soil layer in mm

n Porosity

Drainage maximum

The drainage maximum is the maximum of water a green roof can hold with the possibility to drain rainwater at the bottom. The water which stays in the roof is trapped due to capillary forces. The water what drips out comes out due to gravity forces. After every rainfall the water content of a green roof will at least drop under this drainage maximum. The drainage maximum of a green roof is dependent from the type of soil, its permeability and the slope of the roof.

Plant minimum

The last boundary limit is at which water content level a plant starts to wither. The water content level of a green roof should never be below this level, because else the plants will start to die. The plant minimum is dependent for the type of plant used.

Typical water content, earth and agriculture science

From earth and agriculture science typical ranges for water contents are available. (Rawls W.J, 1982) investigated several typical water contents in America. Typical data are described in the table below. One remark is that in earth and agriculture science the field capacity is measured.

This is a drainage maximum for soil three days after a rainfall. The drainage maximum for a green roof is measured right after a rainfall and will always be slightly higher.

Name Typical water content Description

Volumetric maximum 0.35 to 0.5 * Thickness Fully saturated water, eq to effective porosity Drainage maximum 0.1 to 0.4 * Thickness Soil moisture at 0.33 Pa tension

Plant minimum 0.01 to 0.25 * Thickness Minimum soil moisture at which a plant wilts

Figure 8 Table typical water contents of soil

13 Bachelor Assignment Boundary capacities

The boundary limits are the ends of the boundary capacities of water content in green roofs.

With the limits boundary capacities can easily be determined. The boundary capacities are water contents in liters per square meter or, equally, mm of water.

Figure 9 Water content Boundary conditions

Drain capacity

Total capacity of green roof to drain (and delay) rainwater. Water will drain from the roof if gravity forces are stronger than capillary forces. The drainage maximum is reached when these forces become equal. In nature green roofs will never reach their volumetric maximum. This is depended from the intensity of a rain event. This means that not the whole theoretical drain capacity is used and therefore an effective drain capacity is introduced in chapter 2.4, where the drainage part is discussed in detail.

Retention capacity

No more water will drain from a green roof if after a rain event the drainage maximum water content level is reached. After that the evapotranspiration starts. Water will evaporate and makes space for water to be retained in future rain events. This evapotranspiration process can take several days. However irrigation strategies have a big impact on how much space can be cleared for retention of future rain events.

Bottom capacity

At a certain water content plants are not able to extract water and will start to wilt. Capillary forces are too big for plants to extract water.

Water content Boundary conditions

Volumetric maximum

---

Drainage maximum

---

Plant minimum

---

--- Drain capacity

--- Retention capacity

--- Bottom capacity

14 Bachelor Assignment

Research Results: Water content Boundary conditions

Actual research has been done on extensive green roofs. As mentioned in the research decisions two influencing factors are examined in detail. The soil thickness and the roof slope have been examined at the far ends of their boundaries. In consultation with the architect was decided that extensive green roofs build in the area of San Pedro Sula will probably have soil thicknesses between 2 and 3 inches. However in the report the metric system will be used and therefore the thicknesses are converted to 5.1 cm and 7.6 cm. The roof slopes will have dimensions somewhere between 2 and 20 degrees. This is because of discharge and erosion reasons. In total this makes that the research boundaries of the two influencing factors are:

Soil thickness 5.1 cm & 7.6 cm

Roof slope 2 degree & 20 degree

The research results on the boundary conditions: the volumetric maximum, the drainage maximum and the plant minimum are presented in the table below. Also the drain capacity, Retention capacity and Bottom capacity are calculated. The results are presented in liters per square meter and this equals mm of precipitation. On the next page a graph is published for better understanding.

7,6 cm soil - 2 degree slope 5.1 cm soil - 2 degree slope

Wc

%

Boundary ( l/m2)

Capacity ( l/m2)

Wc

%

Boundary ( l/m2)

Capacity ( l/m2)

Volumetric maximum 43% 32,6 Volumetric maximum 43% 23,9

Drain capacity 8,6 Drain capacity 6,3

Drainage maximum 32% 24,0 Drainage maximum 32% 17,6

Retention capacity 22,5 Retention capacity 16,5

Wilting minimum 2% 1,5 Wilting minimum 2% 1,1

Bottom capacity 1,5 Bottom capacity 1,1

7.6 cm soil - 20 degree slope 5,1 cm soil - 20 degree slope

Wc

%

Boundary ( l/m2)

Capacity ( l/m2)

Wc

%

Boundary ( l/m2)

Capacity ( l/m2)

Volumetric maximum 43% 32,6 Volumetric maximum 43% 23,9

Drain capacity 12,3 Drain capacity 9,0

Drainage maximum 27% 20,3 Drainage maximum 27% 14,9

Retention capacity 18,8 Retention capacity 13,8

Wilting minimum 2% 1,5 Wilting minimum 2% 1,1

Bottom capacity 1,5 Bottom capacity 1,1

Figure 10 Table: Test setup results boundary conditions

15 Bachelor Assignment

The volumetric maximum has been determined with a full saturated test which determines the porosity. The porosity of the tested green roofs is determined to be 0.43. Information on this test is found in the data report. With the total thickness of the soil layer (5.1 and 7.6 cm) the volumetric maximum is determined.

The drainage maximum is determined in different tests on evaporation and drainage rates.

Research data can be found in the data report. The drainage maximum is a rough estimation and is different for sloped roofs.

The wilting minimum is determined with literature and is taken to be 2%. This is because to measure this wilting minimum a part of the green roof has to be taken apart and this would influence further measurements. Special equipment was not available. The wilting minimum has not a big influence on the water handling capacities, because it is a good assumption that in general a green roofs will not reach this point due to irrigation.

The determined water content boundary conditions for the four test roofs are presented in the graph:

Figure 11 Graph: Test setup results boundary conditions

It is clear that the 7.6 cm soils have a higher volumetric maximum. Thereby they have a higher retention capacity and drain capacity as the 5.1 cm soils. When a roof is placed under a slope the retention capacity drops a bit. This is because there is a slightly higher head of water. For sloped roofs is in theory the drain capacity higher, but the effective used drain capacity is lower.

This is explained in paragraph 2.4.

1,5 1,5 1,1 1,1

22,5 18,8

16,5 13,8

8,6 12,3

6,3 9,0

0,0 5,0 10,0 15,0 20,0 25,0 30,0 35,0

7,6 cm soil - 2 degree slope

7.6 cm soil - 20 degree slope

5.1 cm soil - 2 degree slope

5,1 cm soil - 20 degree slope

Water content Boundary conditions test roofs

Drain capacity Retention capacity Bottom capacity

wat er co n te n t (liters/ m 2 o r m m )

16 Bachelor Assignment Influence of design parameters Soil thickness

The soil thickness has a direct influence on the boundary conditions. An increased soil thickness increases the volumetric maximum (Eq. 7) and thereby increases the maximum drain capacity and the maximum retention capacity as well.

Soil buildup

The soil build up, and mainly its porosity has a direct influence on the boundary conditions. However the architect uses always a soil mixture with sand and gravel. There will be not a big difference in the porosity. The porosities of soil is general between 0.35 and 0.5. (Rawls W.J, 1982) Therefore the influence of the porosity is quite limited.

Plant type

The choice of plant type will have a small influence on the boundary conditions of green roofs. Different plants have different wilting boundaries (water content needed to survive.)

Slope

The slope of the roof has a small influence on the boundary conditions of the roof. A roof

with a higher slope (examined 20 degree) has a little higher water pressure and thereby

the roof is able to hold a bit less water than an almost flat roof. The volumetric

maximum in comparison is the same. But the retention capacity for a sloped roof is

lower.

17 Bachelor Assignment Conclusion boundary conditions

For extensive green roofs with soil layers up to 7.6 cm of soil the boundary conditions have an upper limit of about 32 liters water per square meter. This volumetric maximum decreases with lower soil thickness and lower porosities. The maximum water content is divided into three areas: the drainage capacity, the retention capacity and the bottom capacity. The amounts for the four examined green roofs are presented in the following graph:

Figure 12 Graph: Test setup results boundary conditions

A part of the retention capacity will be used for the retention of rain water due to evapotranspiration. This is explained in paragraph 2.3. A part of the drainage capacity will be used for a delay in discharge to the sewer system. This is explained in the drainage paragraph 2.4.

1,5 1,5 1,1 1,1

22,5 18,8

16,5 13,8

8,6 12,3

6,3 9,0

0,0 5,0 10,0 15,0 20,0 25,0 30,0 35,0

7,6 cm soil - 2 degree slope

7.6 cm soil - 20 degree slope

5.1 cm soil - 2 degree slope

5,1 cm soil - 20 degree slope

Water content Boundary conditions test roofs

Drain capacity Retention capacity Bottom capacity

wat er co n te n t (liters/ m 2 o r m m )

18 Bachelor Assignment 3.4 Retention

Retention equals evapotranspiration

After a rain event the water content level is assumed to be at the drainage maximum. After that the water starts evaporating until the next rain event. The capacity to retain rainwater is determined by the water which is evaporated in time. This because the water content balances (eq. 4 & eq. 5) from chapter 2.1 have to be equal.

Evapotranspiration model of Hargreaves (long term)

A popular equation to describe evapotranspiration (evaporation + transpiration) is the Penman- Monteith equation. (ASCE, 1996) However to use the penman-monteith equation detailed temperature data are required. Another simplified version of the penman-monteith equation is the equation of Hargreaves. It uses daily average temperatures to describe a model for the evapotranspiration. This empirical model is a normal used model to predict evapotranspiration.

(Allen, 2006) The equation of Hargreaves looks as follow: (ASCE, 1996)

0

ET   K ET

(eq. 8)

In this equation the evapotranspiration is described by a reference evapotranspiration ET

_and

a crop coefficient K

which is depended from the used type of vegetation.

The reference evapotranspiration is described by the following equation:

0.5

0

0.0023 (

) (

17.8)(

) ET   T  T  T  R

(Eq. 9)

With:

T

Maximum daily temperature T

Minimum daily temperature T

Mean daily temperature R

Sun radiation in MJ/m

/d

In this equation the reference evapotranspiration is calculated by determining the total amount

of incoming sun radiation R

and the influence of temperature differences during day time.

19 Bachelor Assignment

The total amount of sun radiation is different for every other location in the world. The total amount of sun radiation in MJ/m

/d is calculated with the following equation:

 

24 60

sin( ) sin( ) cos( ) cos( ) sin( )

A k r s s

R G d      



  

(Eq. 10)

With:

(283 ) 0.4093sin 2

365

  ^   ^ J ^ 

  (Eq. 11)

&

1 0.033cos 2

r

365

d       J    _{(Eq. 12)}

&

 

arccos tan( ) tan( )



   

(Eq. 13)

With:

R

Sun radiation in MJ /m

/d G

Solar constant (4.92 MJ/m

/d) d

Inverse distance factor earth-sun



Sunset hour angle in radians

 Latitude in radians

 Solar declination in radians

J Day of the year

20 Bachelor Assignment

For the city of San Pedro Sula (Latitude: 15.27°) the sun radiation during a year is presented in the following graph:

Figure 13 Graph: annual sun radiation SPS

With these data of sun radiation and the maximum and minimum temperature differences the reference evapotranspiration can be determined by using equation 9.

Figure 14 Graph: Annual maximum Evapotranspiration SPS

Minimum and maximum temperatures are used from a worldwide weather station, Allmetsat, to determine the Maximum reference evapotranspiration. (Allmetsat, 2009) Data can be found in the data report.

0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00

Sun radiation (in MJ/m2/d)

Sun radiation

Ra - Sun radiation

0,00 2,00 4,00 6,00 8,00 10,00 12,00

Reference evaporation (mm/d)

Max ET

Max ET0

21 Bachelor Assignment

The crop coefficient K

in the formula of Hargreaves is empirically determined for lots of different plant types. A few are noted in the following table. More K

coefficients can be found in the hydrology handbook for civil engineers, page 186.

Description Kc1 Kc2

Grass pasture, rotation 0,40 0,85 Grass Pasture, poor manage 0,30 0,75

Open water 1,05 1,05

Figure 15 Crop coefficients

K

=General factor under unfrequented soil wetting K

=Sub humid climate, average wind speed

For grass this factor is between 0.30 in the early stage of growth and with fully grown condition the K

can be as high as 0.75. These lines are drawn in the graph.

Following the equation the evaporation rates are between 2 and 8 liters per square meter per day for grass as vegetation, dependent on the day of the year and the development of the vegetation.

Figure 16 Graph Evapotranspiration SPS with crop coefficient factors 0,00

2,00 4,00 6,00 8,00 10,00 12,00

Averegae evapotranspiration in mm/d

Evapotranspiration in time

Max ET0 Kc=0,30 Kc=0.75

22 Bachelor Assignment

Research Results Evapotranspiration (long term)

During a month measurements have been taken on the evaporation rates of the 5.1 and 7.6 cm test roofs. The roofs where placed in two sets with a two degree slope.

After a certain amount of time the retention of rainwater was measured. This was done by fully saturating the roof and measure the difference in irrigation and run off. From equation 5 this equals the evaporation rate in time.

Figure 17 Average retention test setup

A trend line is added and so the average evapotranspiration rate for the two degree roofs can be determined.

7.6 cm soil 4.3 liters per square meter per day 5.1 cm soil 3.9 liters per square meter per day

The difference between the 2 and 3 inch roofs is due to lower evaporation rates on a daily bases when there is a low amount of water present in the roof. This effect is explained in the next paragraph.

The evaporation rates were determined during the 18 May to 17 June with average day temperatures between 30 and 35 degrees. During these days there was no actual rainfall. All the additional water was provided by the researcher himself.

y = 4,3425x R² = 0,9974

y = 3,9196x R² = 0,9992

0 20 40 60 80 100 120 140

0 10 20 30 40

Amount of retention in ml/m2

Time in days

Average retention (month)

7.6 cm roof 5.1 cm roof

Lineair (7.6 cm roof) Lineair (5.1 cm roof)

23 Bachelor Assignment

Combining the evaporation theory of Hargreaves with the actual measured evaporation rates delivers the next graph.

Figure 18 Evaporation rater Theory of Hargreaves combined with test setup results

The theoretical rates seem to be a bit high, although the green roofs where starting in their growing phase and suffered some growing difficulties due to some tests under wilting conditions.

0,00 2,00 4,00 6,00 8,00 10,00 12,00

Evaporation in mm/day

Evaporation rates

Max ET0 Kc=0,30 Kc=0.75 7.6 cm roof 5.1 cm roof

24 Bachelor Assignment

Evapotranspiration model (short term)

Another model for describing evapotranspiration rates on a short timescale is discussed in the hydrology handbook for civil engineers. This evaporation model gives a closer look on the evaporation on a short time scale. (ASCE, 1996)

The model explains that water on soil evaporates in two stages. The first stage is water which stays after a rain event on top of the soil layer or in the leaves of vegetation. This water is evaporated very quickly because there are no capilary or molucalair forces to hold the water. In stage two water starts to evaporate at a certain rate. When less water is available for evaporation the capilary and moleculaire forces hold the water much stronger so the evaporationrate drops in time.

Figure 19 Short time evapotranspiration model

25 Bachelor Assignment

Research Results Evapotranspiration (short term)

This short term evaporation effect is examined for the 7.6 and 5.1 cm roofs under a 2.0 degree slope.

Figure 20 Evapotranspiration after x-days without precipitation

The results are presented as the possible amount of retention after a certain amount of days.

The evaporation rate discussed on the previous page is the curve of the line. The first stage evaporation could not be identified. After the first few hours the line is quite steep and therefore the evapotranspiration is high. After a few days the line gets less steep and therefore the evaporation rate drops in time.

It is shown that in the beginning the rates for the 5.1 and 7.6 cm roofs are more or less the same. After day three the two inch roof starts to have more trouble evaporating water, so the evaporation rate drops. After day five the plants start to whither and the 5.1 cm roof reaches its wilting boundary condition. (Chapter 2.2) For the 6.7 inch roofs it takes about a day longer to evapotranspirate the whole retention capacity.

The mean evaporation rate for the first three days is more or less 5.0 liters per square meter per day. This is a bit higher than the evaporation rate with the long term condition due to the fact that the long term evaporation measurements include some data of evaporation after 3 to 6 days without precipitation.

The data of this research part are added in the data report.

0 5 10 15 20 25

0 1 2 3 4 5 6

Amount of retention in l/m2

Time in days

ET after x-days without precipitation

7.6 cm roof 5.2 cm roof

Max retention capacity Max retention capacity

26 Bachelor Assignment Irrigation strategy

The irrigation strategy is a very important factor in the determination of the capability of green roofs to retain rainwater.

Due to high temperatures and long periods without precipitation all the build green roofs in San Pedro Sula should have an irrigation system.

Three different types of irrigation strategies can be recognized.

1. Aggressive Every day without precipitation irrigation takes place.

2. Adaptive Only when it is necessary for plants irrigation takes place.

3. None Poorly maintained roof, plants will die.

At the research site a number of green roofs is installed and they have irrigation systems installed. The irrigations strategy to maintain the green roofs is aggressive. Irrigation takes place every day without precipitation. This makes that the green roofs always have a lot of water in them and the water content always is close to the drainage maximum. A red block is added in the evaporation/retention graph presented in the previous chapter to show in which area the evaporation rates (and thereby retention capacity) will be. This is about 5 mm of water after one day.

Figure 21 Evapotranspiration after x-days without precipitation with irrigation strategy

Also if precipitation comes day after day the capacity of green roofs to retain water will just act like it is irrigated every day and thereby has a retention capacity of more or less 5 liters per square meter per day as well.

With an adaptive irrigation strategy green roofs could have a higher capacity of retention, but it

could prove to be more difficult for plants to grow. The none irrigation strategy is taken to not

take place. Effects will be kind of the same as with an adaptive strategy, but it is not likely that

plants will survive during dry periods.

27 Bachelor Assignment

Other influencing parameters Evapotranspiration Soil thickness

An increased soil thickness increases the retention capacity and so it will take longer to evapotranspirated the whole retention capacity and thereby longer for plants to reach the wilting point. An increased soil thickness has not directly an influence on better retention capabilities. This is because it only increases the time to evapotranspirate all the water, but not the evapotranspiration rate. A thicker soil layer could make that it is not necessary to have an aggressive irrigation strategy and use a more adaptive strategy. This could indirectly increase the water retaining capacities.

Plant type

The choice of plant type has a direct influence on the evapotranspiration rate. This is explained with the formula of Hargreaves. Investigations by the ASCE give different values of K

-factors. A small list is added in Appendix B. In general plants with large surface areas of leaves make a plant transpirate more water, but will give more shadow to the soil, so soil evaporation will decrease. (ASCE, 1996)

Slope

The slope has an influence on the evapotranspiration rate in combination with the orientation.

Orientation

The orientation of a green roof, north, east, south, west has an influence on the evapotranspiration rate due to the fact that more or less sun radiation is captured by the roof.

This only has an effect on sloped roofs; flat roofs do not really have an orientation. In area’s around the equator the sun goes during the day from east to west in a perfect line and has not really a north or south orientation (small seasonable effect) Roofs directed to the north and south take thereby full sun al day. (Test setup configuration) Orientations to the east and west have less sunlight in the evening or respectively in the morning. This has an effect on the evaporation rate. A research done in Belgium shows that the influence of the orientation could be substantial, but they conclude that more research is necessary to give valid results.

(Mentens, et al., 2003) Year period

The period of the year has an influence on the evapotranspiration capabilities of green roofs. On

countries around the equator the amount of sun radiation during a year does not change that

much. The reliability of green roof in equator areas will there for be higher than in more

northern or southern areas. During the year period there are (sometimes large) differences in

amounts of precipitation. Also rain events quick after each other have an influence on the time

to evaporate water in between and thereby on the retention capabilities.

28 Bachelor Assignment Retention conclusion

Retention capacities of green roofs are highly dependent of the evaporation rate and the irrigation strategy which is used. Research results show that the evaporation rate for the grass covered green roofs is about 5 liters per square meter per day in the first three days after a precipitation. This is equal for as well the green roofs with a soil thickness of 7.6 and 5.1 cm.

Green roofs with other plant types will have different evapotranspiration rates, but they will be in same order of amounts.

Because of a frequently used aggressive irrigations strategy which waters the green roofs to

their drainage maximum every day without precipitation the expected retaining capacity is

equal to be the same as the evaporation rate on one day. This is about 5 liters per square meter

per day and equals a precipitation of 5 mm.

29 Bachelor Assignment 3.5 Run off & drainage

During a rain event runoff water leaves the roof in two ways, Direct Runoff and Drainage. In most of the research done on green roofs no differential is made between the two types. This is because it is hard to measure the difference between the actual runoff and drainage, but there are some things to say about the runoff and drainage.

- First, direct runoff of water is more likely to cause erosion on the top layer of the soil.

- Second, drainage of rainwater will extend the period of time of water drainage to the sewer system and could thereby be preferable.

The reason why normally no distinction is made between the direct runoff and drainage is the high amount of factors which influence the distinction between the two types. Factors which can be examined are the slope, type of soil, type of plant and the thickness of the soil, but other parameters like the spread of plants on a green roof and the appearance of little drain paths inside a green roof will influence the distinction that much, that it is impossible to make up a full analytical model. With existing theories of drainage through soils it is possible to give some boundaries for the amount of drainage of rainwater after a rainfall or peak rain event.

*An extra model on the forecasting of runoff, the curvenumber method is added in appendix B.

This is for the more experienced reader.

Drain capacity

The maximum amount of water which can be drained and leave the roof after a rain event or heavy precipitation peak is maximized by the difference between the volumetric maximum and the drainage maximum per square meter. (chapter 3.1) This is called the drain capacity. The drain capacity in liters per square meter or mm precipitation for the 5.1 and 7.6cm soils are:

7.6 cm soil 5.1 cm soil

2 degree slope 8.6 6.3

20 degree slope 12.3 9.0

Figure 22 Drain capacity green roofs

However this drainage capacity is never fully used because in nature the roof will drain water during a rain event. The effective used drain capacity is simulated with heavy precipitation rates by the researcher. For the 2 degree roofs the simulated maximum is about 80 to 90% percent of the total drain volume. For the 20 degree roofs this is only about 40% (because direct run off start to appear). The effective volumetric drain capacity is presented in the next table in (l/m2):

7.6 cm soil 5.1 cm soil

Eff. 2 degree slope 7.0 6.3

Eff. 20 degree slope 4.0 3.3

Figure 23 Effective drain capacity green roofs